Seasonal dynamics in the trophic status of afer (Günther, 1868) () in a Nigerian rainforest stream

Richard P. KING (1)

The stomach contents of comprised primarily midwater invertebrates, allochthonous macro- phyte debris, allochfhonous invertebrates and preyfish. Epiphytic algae, benthic inuertebrates, neustonic invertebrates and miscellaneous insects were secondary components while zooplankton mas of incidental importance. The fish thus utilized food from the three major spafial levels of the stream surface, midwaier and botfom. The diet was dominated by occasional food-types which were consumed during tèss than 3 months of the year. It was closely linked 10 the distributional ecology of the fish in ihe vegetated marginal water biotope where most of the dietaries abound. There was considerable seasonal plasticity in the specific food-types utilized and this resulted in high year-round trophic divers- ity. A wide trophic spectrum and attendant expanded diet breadth were considered as attributes of high foraging success of P. afer.

KEYWORDS: Papyrocranus afer - Nigeria - Seasonality - Dietary habit - Fish.

RÉSUMÉ

L’ÉVOLUTION SAISONNIÈRE DU RÉGIME ALIMENTAIRE DE PAPYROCRANUS AFER (GÜNTHER, 1868) (NOTOPTERIDAE) DANS UNE RIVIERE FORESTIÈRE DU NIGERIA Le contenu stomacal de Papyrocranus afer est constitué d’invertébrés de pleine eau, de débris de plantes el d’insectes allochtones, et de poissons. Les algues épiphytes, les invertébrés benthiques et divers insectes sont des composantes secondaires tandis que le zooplancton n’est qu’occasionnel. Il apparaît ainsi que ce poisson utilise les trois principales zones de son milieu : la surface, la pleine eau ef le fond. Le régime es1 principalemenf opportuniste avec des proies qui sont consomn@es durant moins de trois mois par an. Il est également en relation étroite avec la répartition du poisson dans les zones herbeuses de bordure. On a observé une grande plasticité du régime qui se traduit par une diversité importante sur un cycle annuel. Ce large spectre trophique est considéré comme caractéristique de l’espèce.

?V~OTSCLÉS : Papyrocranus afer - Rivières’- Nigeria - Régime alimentaire - Poissons.

(1) Deparlmenf of Zoology, University of Uyo, P.M.B. IOl7, Uyo, Akwa Iborn Siate, Nigeria.

Rev. Hydrobiol. trop. 27 (2) : 143-155 (1994). 144 R. P. KING

INTRODUCTION southeastern Nigeria (Fig. 1). The stream has a main-channel total length of 53.5 km between its The quality and quantity of available natural food source. in Ikono Local Government Area and where resources influence aspects of t,he life history of fishes it discharges into the Cross River, close to Nwaniba. including inter alia, growth rate, longevity, repro- Ikpa River has a watershed area of 516 km2, of rluc.tive investment,, sexual maturity and fecundity. which 76 km” (14.8 %) of the lower reaches is liable Diet diversity is a measure of t.he degree of complex- t,o annual flooding of the fringing low land riparian ity or specialization in it.ems consumed. Some fish zone during the rainy season. The nonflood zone of species are oligophagous, utilizing a limited number the Upper reac.hes has a basin area of 440 km2 of food resources while others are polyphagous, sub- (85.2 %) and mean dept,h and width of 2.0 m and sisting on wider spect,ra of items. 12.5 m respectively. The adaptat.ive significance of a broad trophic The ent.ire length of the main channel of the spectrum (high diet diversity) is that it ensures a stream lies at the interface of two different geolog- const,ant# energy source, facilitates adequate utiliza- ical deposits: tertiary sediment.ary rocks and cretace- tion of available food resources and enables the fish ous deposits (see geological map of Nigeria in CLAU- to easily swit.ch from one food source to another in SEN, 1964). The stream is considerably shaded by response to nat,ural pulses in their relative abund- overhanging canopy of riparian vegetation (mostly ances and availability (KING, 1993). Conversely, Elaeis guineensis, Raphia hookeri, Raphia vinifera oligophagous fishes (low diet. diversit.y) would be ad- and other tropical forest trees). The aquatic macro- versely affected should there be a dwindle or collapse phytes are mainly Nymphaea, Vossia and Crinium in t.he food resource-base. species. The seasonal variation in stream surface Tropical riverine fishes inhabit dynamic aquatic temperature is 23.3-28.0 “C, Secchi disc transpar- environments and are thus exposed to seasonal ency 12.5- 101 cm and pH 6.6- 7.7. Conductivity pulses in hydrometeorological attributes. Seasonal- varies between 13.0 and 69.2 PS cm-l. The stream ity in habitat conditions influences fishes mainly current velocity is 3.5 - 6.9 cm s-1. through qualitative and quantitative changes in The climate of the area is typical of tropical rain- available food resourc.es (LO\VE-MCCONNEL~, 1987). fore&; it comprises dry (November - February) and Seasonal variations in dietary status and diversity of wet (March - October) seasons. Meteorological data fishes presumably stem from the inherent dynamic.s from the University of Uyo meteorological station in food resources availability / abundance and/or sited within Ikpa River basin (Fig. 2) indicate the intrinsic changes in foraging ethology e.g. active pre- precipitation and ambient temperature regimes dur- dilection for specific food-types. In any aquatic eco- ing this study. The dry season was characterized by system, t.he evolutionary adaptations should there- prevalence of dry tropical continental winds from fore favour fishes that have evolved trophic the Sahara desert and low mean monthly precipita- strategies that ensure optimum foraging under a sea- tion; peak dry season occurred in January - Febru- sonally available food resource condition. ary. The wet season was typifled by prevalence of Thr present study was aimed at documenting the moist tropical maritime winds from the Atlantic overall food composition and extent of temporal Ocean and high mean monthly rainfall. Annual rain- plasticity (monthly and seasonal patterns) in dietary fa11 was 255.8 cm. The cyclic hydrological regime is status and diversity of Papyrocranus afer in a Niger- typified by high water level and slow rate during the ian rainforest stream. It cent,eri on the influence of rains and vice-versa in the dry season. the tropical dry-wet. season cycle on the seasonality More descriptive information on the Ikpa River regimes of these trophic attributes. Very litt.le has basin is provided by KING (1989) and TEUGELS et al. heen published on the general biology and ecology (1992). of P. afer. Brief and fragmentary accounts of aspects of its reproductive and food habits are contained in ALB~RET (1982) and TEUGELS et al. (1992) respectively. MATERIALS AND METHODS

Monthly samples of Papyrocranus afer were obtained from each of eight stations in Ikpa River STU DY AREA (Fig. 1) for 5 days during mid-month (November, 1988-October, 1989 inclusive). Fishing was con- The present study was conducted in Ikpa River, a ducted by the use of 3 gill-nets with 30 mm, 50 mm small perennial rainforest tributary stream located and 80 mm stretched mesh respectively and 50 sin- West of t.he lower reaches of the Cross River in gle-hook (size No. 2) set-lines baited with earth f&w. Hydrobiol. trop. 27 (2) : lz3-15.5(1994). SEASONAL TROPHIC STATUS OP P. AFER (NIGERIA) 145

. \ I !ï015’N I

0 Area liable to flood

- - - - Limit of Ikpa River basin -.. l Sampling stations Esuk Inwang

Towns

Streams 0 5 IOkm l 1 I

FIG. 1. - Map of Ikap River, showing the sampling stations. Inset : map of Nigeria showing the locat,ion of Ikpa River. Carte de situafion de la rioière Ikpa et des sfations de pêche, avec les zones inondables (hachurèes).

worms. Attemps to capture P. afer with the valued the potentially available food resources to stream basket traps were unsuccessful. Day and night fish- fishes. This is similar to WHYTE'S (1975) taxonomie/ ing were conducted in the open water, littoral zone ecological groupings of the diets of the ilshes of Lake and fringing swamps. Bosumtwi, Ghana. The merits of this system of cate- The specimens were transported to the laboratory gorization of food in flsh dietary studies are dis- where they were measured to the nearest 0.1 cm cussed by BERG (1979). total length (TL). They were eviscerated and the The frequency occurrence of each item as dom- stomachs slit open. The stomach contents of eac.h inant (by volume) (di) and non-dominant (fi) stomach specimen were placed in a Petri dish and aggregates contents were noted. The integrated importance of dispersed with a few drops of water prior to macro- each item was then expressed by the food ponderal scopie and microscopic (with variable magniflcations index (FPI) (KING, 1991): up to x 40) examinations. The contents were sorted, FpI = (ft + 4) identified and categorized according to taxonomie 7 x 100 (1) and microhabitat (i.e. where an item occurred most 2 (fi+&) commonly) criteria based on author’s field notes on i-1

R~U. Hydrobiol. trop. 27 (2) : M-155 (1994). R. P. KING

in n compared months/seasons. The IBD ranges from 0 (for completely different set of dietaries in each month/season) to 100 o/. (for identical sets of dietar- ies in each month/season). An index of diet diversity (F) was computed from the o/O FPI data using the formula (see ALATALO, 1981; GRUNDEL, 1990): -1 F = ( i$lW1- [=PC-!iZa PJ - 1 (4 i-1 1 where Pi = proportion of the diet comprised by re- source type i and n = number of food c.at.egories in t.he diet. This index is sensitive to changes in two attributes viz food richness and equitability (= the 60 - degree to which a11items are equally represented); it 50 - increases as food richness and it,s equit.ability increase and declines when few items dominate the 40 - E diet. The F is scaled such that 1 .O represents an even ; 30. distribution and zero, a strongly skewed distribu- E 20. tion. c? TO evaluate short-term (i.e. month-to-month) and 10 - long-term (i.e. seasonal) changes in diet composition, o- - the percentage similarity coefficient S was used Nov. Dec. Jan. Feb. Ma dbr. Apr. May Jun. Jul. Aug. Sep. Oct (Moss and EATON, 1966):

FIG. ‘2. - Srasonal variation in some meteorological S = j) min (Xi,Yi) (5) parameters of Ikpa River basin at. Uyo. i=1 Lwariafions climatiques à i!yo, dans le bassin de la rioière Ikpa. where Xi and Yi = proport.ions of the components of the series of nth comprising the diets of X and Y. This index ranges from zero, for total17 disimilar dietary compositions, to 100 O& for identlcal diets. This index has a range of O-100 7,; items wit.h The significance of the deviat.ions of calculated FPI 2 10 yo were arbitrarily considered as primary values of S and IBD from maximum possible values dietaries, thosp with FPI = 1.0 - 9.9 yo as second- (i.e. S,,, = 100 yo and IBD,,, = 100 yo respect.- ary and those with FPI < 1.0 yo as incidental. The ively) were evaluated by X%est (BAILEY, 1959). percent, compositions were used to describe t.he over- The magnitude of variat,ion in F was assessed by the a11diet and temporal changes in food habit,. coefficient of variation (LOWETIN, 1966), regarding Thr regularity with which each food-type was con- values above 30 o/Oas high. sumed on a monthly basis was evaluated by a per- centage constancy index (CI): cL = mi loo (2) LZI RESULTS where m; = number of months in which food-t.ype i was consumed; M = t.otal number of months in t.he Papyrocranus afer was available a11year-round in study period (here 12 months). This index is scaled catches from the eight sampling stations, thus indic- from 0 (for absence throughout the year) to 100 yo ating that it is a resident, species in t.he stream. It (for occurrence in every mont,h of the year). The was found to be a solitary limnophilic fish that was KOCH’S (1957) composite index of biotal dispersit.y most c.ommonly caught in quiet and gently-flowing (IBD) was used to assess how widely dispersed the vegetated littoral zone and fringing swamps of the dietaries were hetween months and seasons: stream. It tended to evade Swift open water IBU = (T - S) 100 C3) stretches of the stream. There was no preference for any particular substrate-type as it was caught, over qn- 1) both muddy and sandy bottoms; t,here were also no where T = arit.hmetical sum of dietaries in each of n temporal and ontogenetic regimes in habitat prefer- compared months/seasons; S = t.otal list. of dietaries ence.

Rw. Hydrohiol. trop. 27 (2) : 14&1% (1994) SEASONAL TROPHIC STATU~ OF P. AFER (NIGERIA) 147

TABLE 1 water invertebrates (35.8 %), allochthonous macro- phyte debris (16.0 Oh), allocht.honous invertebrates Fuionthly numbers and sizes of P. afer from Ilrpa River examined for food (November 1988 - October 1989) (14.2 %) and preyflsh (12.5 %); sand grains (7.8 OA), Nombre et taille des spécimens de P. afer capturés et examinés epiphytic algae (4.6 O/& miscellaneous insects (novembre 196X-octobre 1989) (3.5 %), bent,hic invertebrates (3.4 %) and neustonic invert,ebrates (1.5 %) were secondary dietaries while zooplankton (0.7 %) was of incidental importance. Months Nurnlw of specimens Total length Of the epiphytic algae browsed, filamentous forms examined b-d (Spirogyra, Ulothrix, Oscillatoria, Lyngbya) were most important while only small proport.ions of uni- cellular desmids (Closterium, Cosmarium, Docidium, November 34 19.4 - 39.8 Penium, Euastrum, Plurotaenium, Scenedesmus) and December 24 15.1 - 34.0 diatoms (Amphora, Surirella, Pinnularia, Gompho- Jauary 14 15.5 - 36.8 February 7 19.3 - 36.0 nema, Eunotia, Nitzchia) were represented. Coarse March 10 14.2 - 38.7 detrit.us dominated t,he allochthonous macrophyte April 3 18.5 - 35.6 debris dietary comp0nent.s as a primary item May 12 18.0 - 42.0 whereas fine detritus were of secondary imp0rtanc.e. 30 21.7 - 38.0 The detrital particles were derived largely from ;; 20 19.6 - 41.0 decaying shed leaves of overhanging riparian macro- August 16 17.5 - 36.9 Septentber 32 18.7 - 37.5 phytes. Other components of allochthonous macro- October 17 13.9 - 37.6 phyte debris included floating seeds, fruits, flowers, leaf fragments and other vascular plant matter of riparian vegetation. Although a variety of bent,hic invertebrates were utilized (10 categories), each category was of Altogether, 219 specimens of Papyrocranus afer incidental status and constituted less than 0.8 oh of were examined for t,rophic attributes. The number of the diet. It. is however noteworthy that crabs (Pota- spec.imens examined per month and their size limits monautes paecilei) and shrimps (Caridina africana, are presented in Table 1. Although relatively few Macrobrachium dux, Macrobrachium vollenhoveni) in specimens were available during February - April, this diet category are fairly large-sized it.ems when the results indicate that these did not overly influ- compared with the sizes of other benthic inverteb- ence the values of the t.rophic parameters. rates such as ostracods, arachnids and insect.s. The 9 dietary components of midwater invertebrates Diet composition were dominated by anisopteran (dragon fly) nymphs as the single most important constituent (20.9 yo); The overall food composition of Papyrocranus afer this was followed by adult naucorids (Pelocoris) (Table II) shows the utilization of a wide assortment (9.8 %), zygopteran (damsel fly) nymphs (1.3 YA) and of food resources which were assigned to ten major hydrophilid imagoes (1.4 $4) as secondary dietaries ecological categories: epiphytic algae, allochthnous whereas each of the remaining midwater inverteb- macrophyte debris, bent,hic invertebrates, midwater rates, with less than 1.0 o/. contribution to food c,om- invertebrates, neustonic invertebrates, allochthon- position, was of incidental importance. Field ob- ous invertebrates, miscellaneous insects, zooplank- servations revealed that most of the midwater ton, preyfish and sand grains. The food habit was invertebrates were closely associated with submerg- overwhelmingly carnivorous, with 71.5 o/Oof the diet. ent marginal aquatic veget.ation; they were eit.her comprising faunal food-types while items of plant periphytic and/or nektonic among t.he macrophyt,es. origin encompassed 20.7 %. Abiogenic materials The two insect taxa constituting the neu&onic (Sand grains) constitut,ed a small proportion of the fauna weré coleopterans, with gyrinids (whirligig stomach contents. The carnivorous habit of P. afer beatles) as secondary items while t.he velliids (broad- centered predominantly on arthropods which ac- shouldered water striders) were incidentally con- counted for 59.1 o/Oof the diet. The arthropod prey sumed. Allochthonous fauna in the diet comprised comprised various development,al st.ages, including invertebrates that dropped into the wat.er from the imagoes (adults), subadults, nymphs, deutonymphs emergent littoral and overhanging riparian macro- and larvae. phytes as well as t.hose washed into the stream by Of the 45 items that formed the diet, 14 (31 %) surface runoff. The 11 cat,egories of allochthonous were of allocht.honous origin while 31 (69 %) were of invertebrate dietaries were dominated by adult for- autochthonous sources. Primary dietaries were mid- micids (black ants) (4.2 yo) and coleopterans (4.7 %).

R~U. Hydrobiol. trop. 27 (2) : 143-X5 (1994). Overall and seasonal trophic spectra, and constancy indices of t,he dietaries of P. afer in Ikpa River. Spectre trophique global et saisonnier, ef indice de constance (CI) d u régime alimentaire de P. afer dans la rioiére Ikpa.

Food items % Food Ponderal Index @PI) % Constancy index Overall composition Dry season Wet season (CI) Epiphytic algae Filamentous algae 2.3 2.2 2.4 58.3 Unicellular algae - Desmidiaceae 1.6 1.6 1.7 50.0 Baccillanophyceae 0.7 0.7 0.7 33.3

Allochthonous macrophyte debris Coarse detritus 10.5 14.2 8.9 91.7 Fine detritus 4.6 4.0 4.8 71.0 Other debris 0.9 1.5 0.7 33.3

Benthic invertebrates Nematoda (free living) 0.1 0.4 8.3 Ephemeroptera nymphs 0.4 0.7 0.2 16.7 Diptera - Chironomid larvae 0.2 0.4 0.2 16.7 Trichoptera larvae 0.4 1.1 16.7 Crustacea - Ostracoda 0.1 0.2 8.3 Potamidae : Potamonautes puecilei 0.2 0.3 8.3 Atyidae Caridina africnna 0.2 0.4 0.2 16.7 Palaemonidae : Macrobrachium dux 0.5 0.7 0.3 16.7 Mucrobrachium volZenhoveni0.6 1.1 0.3 16.7 Arachnida 0.7 0.4 0.9 41.7

Midwater invertebrates Odonata - Anisoptera nymphs 20.8 20.4 20.9 91.7 Zygoptara nymphs 1.3 0.7 1.6 41.7 Coleoptera - Hydrophilidae adults 1.4 2.9 0.7 50.0 Dytiscid : Adults 0.7 1.0 16.7 larvae 0.4 0.5 16.7 Hemiptera - Naucorid adults 9.8 7.6 10.9 100.0 Nepidae : Nepn adults 0.9 2.2 0.3 33.3 Notonectidae : Notonecta adults 0.2 0.4 8.3 Unidentified Heteroptera adults 0.3 0.7 8.3

Neustonic invertebrates Coleoptera - Gyrinidae 1.3 3.3 0.3 25.0 Velliidae 0.2 0.3 8.3

Allochthonous invertebrates Orthoptera - Gryllidae adults 0.2 0.3 8.3 Acrldidae adults 1.3 0.4 1.7 33.3 Hymenoptera : Formicidae adults 4.2 1.5 5.5 75.0 Isoptera : adult termites (wlnged reproductives) 0.1 0.2 8.3 Lepidoptera larvae 0.5 1.5 8.3 Copeoptera : Adults 4.7 5.4 4.3 66.7 Larvae 0.7 1.5 0.3 25.0 Anisoptera adults 0.1 0.2 8.3 Unidentified terrestrial adult insects 1.8 2.6 1.4 50.0 Arachnida - Aranae (adult spiders) 0.1 0.4 8.3 Diplopoda - Juliformia 0.5 0.7 16.7 Miscellaneous insects 3.5 4.7 2.9 66.7

Zooplankton Crustacea - Cladocera 0.4 0.5 16.7 Copepoda 0.1 0.2 8.3 Rotifera 0.2 0.3 8.3

Preyfish Whole fish 3.6 4.0 3.4 66.7 Fish scales 8.9 5.8 10.4 91.7 Sand grains 7.8 4.7 9.4 100.0

Arachnida = (Leimnohalacarid mites) : Hydrachnella adults and deutonymphs. Reu. Hydrobiol. trop. 27 (2) : 143-k% (1994). SEASONAL TROPHIC STATUS OF P. AFER (NIGERIA) 149

These were accompanied by unidentifled insects (a) constant elements: 14 (31’ o/Oof total number of (1.8 OA) and acridids (short-horned grasshoppers - food-types ingested) items which appeared in the Zonocerus) (1.3 %) while each of the ot,her taxa con- stomachs during 6-12 months of fhe year (i.e. CI 2 stituted less than 0.6 yo of the diet. Imagoes of gryl- 50 %)1 lids (crickets) and isopterans (termites) consumed (b) accessory elements: S (17.8 %) items which were represented by the genera, Acheta and Macro- occurred in t,he stomachs for a period of 3-5 months termes respectively, both of which are known to ex- (CI = 25 - 40 %), hibit high nocturnal activity. The allochthonous lep- (c) occasional elements: 23 (51.1 Oh) food-types idopterans (butterfly larvae), coleopterans (adults wich were consumed during less than 3 months of and larvae), anisopterans (adults), araneids (spiders) the year (CI < 25 %). and juliform millipeds c,ould not be identifled below the operational levels presented in Table II. It is thus apparent that the diet of Papyrocranus The 3 components of zooplankton were minor afer was dominated by occasional elements, compris- dietaries and formed 0.1-0.4 y0 of the diet. Preyfishes ing items which were opportunistically consumed. The low index of biotal dispersity (IBD = 29.1 Oh) consisted of cyprinodonts (Epiplatys sexfasciatus, Aphyosemion gardneri, Aphyosemion splendopleure, indicated considerable month-t.o-month changes in qualitative food composition. Filamentous algae, Aphyosemion bieiftatum), juvenile tilapiine (Tilapia desmids, coarse detritus, fine detritus, anisopteran mariae) and chromidotilapine (Chromidotilapia nymphs, adult hydrophilids, naucorids, formicids, guntheri) cichlids. The fish scales consumed closely unidentified terrestrial insects, whole preyfish and resembled those of adult T. mariae and C. guntheri. fish scales constituted important. food sources for Several stomachs of Papyrocranus afer were found filled with fîsh scales and no other fish remnants of 6-12 months of the year while a11 other food-types whole fish, thus suggesting that the scales were prob- played accessory or occasional roles in the annual ably resped from the bodies of live fishes. food habit of the fish since they featured in the stom- achs during only l-5 months of t.he year. There was considerable monthly variation in the Diet stability relative importance of the food-types which altern- ated about the primary, secondary and incidental The monthly dynamics in the food components of scales (Table III). Epiphytic algae formed secondary Papyrocranus afer revealed that, they cari be arbitra- dietaries during June - December. A marked decrease rily arranged in three groups: in t.he relative consumption of this item occurred as

TABLE III Monthly variations in the y0 food ponderal index (FPI) of P.afer in Ikpa River. l?volution mensuelle de l’indice pondéra1 (FPI) de P. afer.

Months Food items N D JFMAM J JASO

Epiphytic algae 6.8 4.5 - - - - - ‘8.0 6.8 3.6 4.4 5.3 Allochthonous macmphyte debris 20.5 15.3 17.4 33.3 5.3 30.7 7.9 15.8 11.0 15.6 11.1 24.1 Benthic invertebrates 3.9 7.5 7.7 - 5.3 7.7 2.6 4.0 5.5 1.2 0.7 - Midwater invertebrates 32.5 33.7 40.2 41.7 23.7 7.7 44.9 41.3 37.0 42.1 34.0 30.7 Neustonic invertebrates 5.3 3.0 - - - - 5.3 1.6 - - - - Allochthonous invertebrates 12.1 13.6 19.3 4.2 44.7 23.1 13.2 8.8 12.3 - 18.6 20.1 h&cellaneous insects 1.5 9.1 3.9 12.5 - - 5.3 3.9 5.5 7.2 - - Zooplancton - - - _ - - 5.3 - - 2.4 1.5 - Preyfish 11.3 10.6 9.6 - 15.7 23.1 15.8 10.3 12.3 12.1 18.6 10.6 Sand grains 6.1 3.0 1.9 8.3 5.3 7.7 2.6 6.3 9.6 15.8 11.1 9.2

R~U. Hydrobiol. trop. 27 (2) : 143-155 (1994). 150 R. P. KING the months progressed from June to September but TABLE IV increasrd slightly between October and November. Mean diet. diversity bet.ween adjacent months Allochthonous ma&ophyt(e debris was of primary and mon&to-mont.h food overlap of P. afer in Ikpa River. dietary status throughout the year, except. in March Diversité moyenne ef recouvremenf entre mois de la nourrifure when it was of secondary importance; peaks de P. afer. occurred in February, April, June, August and Oct,o- ber - November. Benthic invertebrates were secondary dietaries in Months Mean diet diversity Food overlap a11 mont-hs except in February/October when they 03 (% SI were not consumed and September when they formed incidental component,s. Midwater inverte- brates dominated the diet of Papyrocranus afer in a11 November - December 0.699 56.6 months except during the onset of t.he rains in March December - January 0.738 58.8 and April when they were replaced by allochthonous January - February 0.780 42.7 invertebrates and allocht,honous macrophyte debris February - March 0.@2 26.9 46.9 respect.ively. Midwater invertebrates were primary March - April 0.871 0.769 25.9 dietaries t,hroughout the year except in -4pril when ApriI - May 0.670 57.3 they played a secondary role. Neustonic inverteb- May - June Jae - Jdy . 0.661 68.9 rates were secondary dietaries in November - Dec.em- July - August 0.635 75.4 ber and May - June. Allochthonous inverteb- August - September 0.669 55.4 rat.es constitut.ed primary dietaries in Novem- September - October 0.802 68.1 ber - January, March - May, July and September- Ortober and secondary diet.aries in February and June. Miscellaneous insects were of primary and sec- ondary dietary value in February and May-August/ November - January respectively. Zooplankton was the dry season than during the rains. Conversely, the a secondary dietary in each of the three months relative importance of midwater invertebrates, (May, August, September) that it was consumed. allochthonous invertebrates, zooplankt,on and prey- Preyfish was of primary dietary status in March- fish were slightly higher during the rains than in the December and of secondary importance in January. dry season. Approximately equal proportions of epi- Sand grains were encount,ered in the stomachs year- phytic algae are consumed in both seasons. round and formed primary content,s in August. - Sep- The low biotal dispersity and mean monthly over- tember and secondary c.ontents in October - ,July. lap indices for the dry (IBD = 37.4 %; S = 52.7 %) Overall, the results est,ablished considerable and wet (IBD = 30.9 %; S = 56.8 %) seasons de- month-to-month plasticity in diet composition, with picted low intraseasonal diet stabi1it.y. Similarly, the overlap values (S) ranging between 26.9 y0 (Febru- significant deviations from maximum possible ary - March) and 75.4 y0 (July - August) (Table IV). values, of interseasonal biot.al dispersity (IBD = Moderat-ely high diet. overlaps (S 2 68 %) occurred 60.0 %: X2= 10.00, df= I, P < 0.01) and diet over- in June - July, ,July - August and Sept,ember - Octo- lap (S = 73.6 %: X2 = 4.015, df = 1, P < 0.05) ber whereas other pairwise monthly diet similarities indices portrayed low interseasonal diet stability. revealed lower overlaps (S < 60 o/. m each case). The magnitude of variability in food composition is fur- t.her attest,ed by the low average month-to-month Diet diversity overlap coefficient (S = 53 %). The composite diet. data for the dry and wet sea- The magnitude of the composite trophic diversit.y sons are presented in Table II. Ostracods, Potamo- (F = 0.614) represented 61.4 yo of maximum possible nautes paecillei, adult, and larval dytiscids, adult diversity (i.e. F,,, = 1.OOO); intra-annual variability veliids, isopterans and anisopterans, juliform milli- was 10~ (CV = 13.1 %). Minimum and maximum peds, cladocerans, copepods and rotifers were not t,rophic diversity were attained in September c.onsumed during the dry season whereas free living (F = 0.576) and March (F = 0.917) respectively. It nematodes. trichopteran larvae, adult not.onectids increased from October to Marc.h and declined from and unidentified heteropterans, larval lepidopterans April to September. Diet diversity index was not and aranae were excluded from the wet season diet. markedly different in the dry and wet seasons Papyrocranus afer fed to a larger extent on alloch- (Fig. 3) but intraseasonal variability was higher t,honous macrophyte debris, benthic invertebrates, in the dry (CV = 18.4 %) than wet (CV = 13.4 %) neust.onic invertebrat.es and miscellaneous insect.s in season.

Hw. H!/drobiol. tmp. 27 (2) : 143-1.55 (1994). SEASONAL TROPHIC STATUS OF P.AFER (NIGERIA) 151

1.0 L one food resource to another. Availability and Dry season 0.623 Wet season 0.636 abundance were net det.ermined in t.his study; how- ever, if the diet composition is a reflection of food availability and abundance in the habitat, the pro- port.ions of the various dietaries in the st,omachs cari be considered as surrogate measures. T~US, t,he “occasional status” of t,he diet constancy index for 0.6 i over 50 o/. of the food-types (Table II) and the low 0.5 L index of biotal dispersit); may be advanced as evi-

I l l I I I dences of diet switching m P. afer. N DJ FM/ktl: JASA A wide feeding spectrum also ensures the reduc- FIG. 3. - Seasonal variation in diet diversity of P. afer t.ion of competition for food between conspeaifics and in Ikpa River. congeners oc.cupying the same marginal water bio- Divers% du régime alimentaire de P. afer dans la rivière Ikpa au tope. The inverse function between diet diversity cours de l’armée. and diet. overlap coefficients (eqn. 6) provides evid- ence that diversificat.ion of diet may reduce intra- specific competition for food. In aquatic systems where the supply of specific components of food-base is unstable or unreliable, the evolution of opport.un- Mean diet diversity index of successive months istic foraging habit as in Papyrocranus afer would and diet overlap coefficients between these months facilitate greater success of the fish than the special- (Table IV) were inversely correlated (r = -0.611, ized foragers. The abundance of P. afer may there- df = 9, P < 0.02) according to the linear function: fore not be limited by food resources but by other F= 0.895-0.0035 (6) ecological factors. Papyrocranus afer is a slow-moving fish that pref- This relationship indicates that diet. diversity erentially inhabits the vegetated littoral zone of decreased as food overlap increased and vice-versa. Ikpa River. Persona1 non-quantitative field observa- tions indicated that. t.his biotope supported large populations of diverse taxa of aquatic invertebrate DISCUSSION fauna as also noted by WELCOMME (1979) and LOWE- MCCONNELL. (1987) for other tropical aquatic sys- The overall stomach c0ntent.s of Papyrocranus afer tems. The’ preferred biotope of P. afer t,hus attunes from Ikpa River depicled a high diet.ary complexity, with the high dietary importance of the midwat.er associated with the exploitation of autochthonous invertebrate assemblage associated with submerged and allochthonous food resources (Table II). macrophytes of marginal Wat#ers. Alt.hough the epi- The trophic st.atus of the fish cari be primarily phyt.ic algae in the diet were proba’bly inadvert.ent.ly assigned to the pelagic (midwater) foraging “general- engulfed along with the aufwuchs fauna, t,hey are of ized invertivore guild” since invertebrates comprised nutritional significance, having a high calorie con- 58.9 y0 of the diet. The fish cari be considered as a tent (ODUM, 1971). WELCOMME (1979) has pointed to “primary consumer” due to the ingestion of algae, the abundance of epiphyt.ic algae as food source on det,ritus and macrophytes; a “secondary consumer” submerged marginal veget,ation in tropical waters. due to the consumption of invertebrates and a “top Other plant food resources of allochthontius origin consumer” because of Ris piscivorous habit. P. afer enter the stream largely t.hrough the marginal wafer is t.hus capable of exploiting the primary, secondary inhabited by P. afer. and tertiary trophic levels of the stream. The prin- Since the presence of submerged vegetation cipal food items cari be ranked in the following order encourages the concentration of benthic invert.eb- of decreasing importance: midwat.er invertebrates < rat,es, the secondary diet,ary role played by benthic allochthonous macrophyte debris < allocht,honous invertebrates for Papyrocranus afer is not astonish- invertebrates < preyfish < epiphytic algae < mis- ing. The neuston, a dietary component of P. afer, is cellaneous insects < benthic invertebrat.es < neu- usually abundant in shelt.ered vegetat.ed areas (WEL- stonic invertebrates < zooplankt.on. COMME, 1979), this c.orresponding with the biotope The selective advantage of a euryphagic trophic preference of the fish. In stream ecosystems, the strategy (heterotrophy) involving dlfferent aquatic input of allochthonous materials (an important food ecosystem energy levels is that it facilit.at,es the source for P. afer) is highest in marginal waters from effective utilization of available food resources and the overhanging macrophytes of t,he riparian zone t,he assoc,iated ability of the fish to easily switch from (KARR et al., 1981). The cyprinodonts and juvenile

Rw. Hytfrobiof. hop. 27 (2) : 113-155 (199$). 152 R. P. KING cichlids consumed by P. afer are known to inhabit ceans. It is energetically more profitable to consume the quiet marginal waters (REED et al., 1967; HOL- large items than small ones (ANGERMEIER, 1985). DEN and REED, 1971). Sand grains may have been inadvertently ingested From the foregoing evidences, it is apparent that during benthic foraging, this assertion being sup- the vegetated marginal water represents a natural ported by the fact that no stomach was found “food concentrate”. Papyrocranus afer thus inhabits exclusively wit-h sand. The presence of associated a biotope in which it is in close proximity to its fine particulate organic matt.er (i.e. organic matter abundant food resources. The trophic bioenergetic. adhering to sand particles) in stomachs indicates signiflcance of t.his food-consumer relat.ionship is that some nutritional beneflt may accrue from the that P. afer prefers a biotope from where its food ingestion of Sand, although this is not anologous to resources are easily accessible and readily obtained that derived by deposite feeders subsisting predom- with limited effort and expenditure of the energy inantly on mud/sand with associated organic matter, c.ost of foraging; consequently, it maximizes the net epipelic/episammic algae and microzoobenthos energy intake and hence, the maintenance of the 6sh (WELCOMME, 1979, 1985, 1986). Moreover, the sand population. grains ingested by Papyrocranus afer may assist in The food habit of Papyrocranus afer in Ikpa River the mechanic.al maceration of food in its muscular- is broadly in accord with the preliminary findings of ized stomach eventhough its st.omach walls are not TEUGELS et al. (1992) and closely parallels the food as muscularized as those of mugilids (BOND, 1979). habit of the allied Xenomystus nigri in Oueme River, The ability of Papyrocranus afer to exploit the sur- Benin Zaire (WELCOMME, 1979) and in equatorial for- face, midwater and bottom water spatio-trophic. est waters of Central Zaire (WELCOMME, 1979; LOWE- niches for allochthonous and autochthonous foods is MCCONNELL, 1987). The predominance of coarse a spect.acular ecological feature which may account det.ritus over fine detritus parallels the dietary habit in part for the wide variety of items consumed, since of Rrienomyrus brachyistius in Ikpa River and is a the flsh is capable of encountering and hence, mak- reflection of the abundance of this item in the stream ing full use of available food resources in the entire (KING, 1989). The ecological signiflcance of this item water column. This phenomenon also ensures con- as principal energy source in headwater streams ha.s stant encounters with food resources regardless of been discussed by Wmxonmm (1979, 1985). The the vetical level at which the fish flnds itself in the overall food composition of P. afer comprised 20.6 y0 stream. plant materials which also featured in t,he diet year- The diet constancy index portrayed that the food round. Although the assimilation effmiency of plant of Papyrocranus afer was dominat,ed by occasional matter by fish is generally lower than that of elements. This points to the high degree opportun- prey, it.s consumption provides satiation by virtue of istic foraging strategy of the llsh. Allochthonous mere filling of the stomach as well as increasing gast- foods such as coarse detritus, fine detritus, macro- rit evacuation rates and henc.e ration size, both of phyte debris, acridids, formicids, coleopteran ima- which compensate for the low nutritional quality goes and larvae and unidentified insect taxa were of (MAGALHAES, 1992). constant or accessory statuses. This type of trophic, Most of the stomach contents of Papyrocranus afer reliance on allochthonously produced food resources were indicative of vigorous reliance on surface, would profitably persist only in stream ecosystems pelagic and benthic foraging. Surface foraging was with thick and stable riparian vegetation that ensure evidenced by the ingestion of surfac.e-dwelling cypri- the regularity. in t,he supply of speciflc. food items nodont preyfish, neuston and allochthonous foods and suggests that large/small scale anthropogenic al- that landed on the water surface. The role of ter- terat,ion in the riparian forest composition and struc- restrial food resources in the trophic. ecology of trop- ture may impose adverse impacts on the trophic sta- ical stream fishes has been noted by several authors, t.us of P. afer. including WELCOMME (1979), ANGERMEIER and The overall euryphagy, low index of biot,al dis- KARR (1983) and LOWE-MCCONNELL (1987). persity and diet constancy index (for most dietaries) Pelagic. foraging was indexed by the consumption of Papyrocranus afer may be relevant in illustrating of midwater fauna (mainly insects and juvenile the selective pressures that lead to the evolution of cichlids) and epiphytic algae while the benthic inver- t,his trophic pat,tern (cf. LOWE-MCCONNELL, 1987). tebrates, sedimented detritus and sand grains in the High trophic specialization is expected of llshes that stomachs at.tested to benthophagy. Although each live in equilibrium habitats and whose dietary habits component, of “benthic. invertebrates” was of incid- are regulated by det,erministic (stable) processes of enta1 import,ance, they c.ould be of muc.h nutri- food supply; alternatively, trophic generalization is tional value by virtue of the large sizes of some of an expected attribute of flshes inhabiting non-equi- them e.g. potamid, atyid and palaemonid c.rusta- librium or stochastic habitats and whose dietary

Rtv. Hytirobiol. hp. 27 (2) : 143-155 (1994). SEA~ONAL TROPHIC STATU~ OF P. AFER (NIGERIA) 153

habits are regulated by the differential responses of and higher rate of dislodgement and downstream the fish to unpredictable or unstable processes of transport of benthic fauna during floods. Reduction food supply. Specialization in dietary habits ean be in the abundance and availability of benthic inver- deterministically regulated in fishes subsistihg tebrates in Ikpa River by one or more of the above largely on autochthonous foods whereas generalized mechanisms, probably accounted for the decline in food habits are stochastically regulated in fishes with their importance during the rains in P. afer heavy reliance on allochthonous foods. This agrees ,(cf. KING, 1989). with the notion of THOnlAS (1964) that t*he availabil- With t,he onset of the dry season and accompany- it,y of allochthonous fauna as fish food is uncertain, ing reduction in stream level and slow rate, the diet- being determined largely by extrinsic fact.ors such as ary importance of benthic invertebrates increased, rainfall, wind, air temperature and the occurrence of perhaps as a consequence of greater abundance and marginal vegetation. availability; this increase in the diet,ary importance In the light of available diet data for Papyrocra- of benthic invertebrates is reflected in the increased nus afer, it cari be infe.rred that the interplay of ingestion of sand grains. deterministic and stochastic processes are likely t.o The dietary importance of midwater invertebrates be critical elements in regulating the trophic st.rat,e- in the wet season slightly exceeded that of the dry gy of the fish. The relative importance of either eco- season but that of neustonic forms was higher during logical factor cari be discerned from the composite of the dry season contrary to LO\VE-MCCONNELL’S food preponderance of autochthonous (66.3 %) and (1975) anecdotal report that the abundance of neu- allochthonous (30.2 %) foods which portrayes determ- stonic fauna increased during the rains. The wet sea- inistic pressures as assuming much greater signific- son increase in the dietary importance of allochtho- ance than st,ochastic pressures in structuring t.he nous invertebrates (cf. LOISELLE, 1971; KING, 1989) dietary strategy of P. afer in Ikpa River. was particularly pronounced in the gryllids, acridids, The low values of the overall index of biotal dis- formicids, isopterans, anisopterans and juliform mil- persity and month-to-month diet overlap coefficients lipeds. The reports of ZARET and RAND (1971), (Table IV) suggest that the euryphagy of Papyrocra- LOWE-MCCONNELL (1975), WELconmm (1979) and nus afer and high degree plasticity in food preference ANGERFJEIER and KARR (1983) lend support to t,he was maintained t,hroughout the year. These are fur- regime observed here (i.e. hightened dietary importe ther buttressed by the high diet diversit,y index in ance of allocht.honous fauna during the rains and most months of the year. Given the supposition that lower st.atus in the dry season). foraging activity of P. afer remained constant Availability of allochthonous invertebrates to throughout the year, the seasona1it.y in the relative tropical stTeam fishes may increase during the rains importance of the dietaries cari be interpreted as a due to higher productivity, expanded stream area crude index of the pulses in food resources availabil- vis-à-vis the dry season, and the mechanical dis- ity and abundance (cf. KING, 1989; MAGALHAES, lodgement and transfer of the fauna from the normal 1992). Thus the dry season increase in t.he relative substrates into the stream through the actions of importance of allochthonous macrophyte debris cor- wind, rain and surface runoff (cf. ANGERMEIER and responds with the period when many forest’ trees KARR, 1983; Pers. observ.). In Ikpa River, t.he shed their leaves. stormy early rains of the March-April seasonal trans- Papyrocranus afer consumed more benthic inver- ition appeared t.o be the period of maximum input of tebrates in the dry season than during the rains in allochthonous invertebrates as indexed by the con- contrast to the reglme in abundance of this food cat- siderable increase in their dietary importance in Papyrocranus afer egory in tropical streams (ANGERRIEIER and KARR, during these months (Table III). 1983). A number of reasons cari be advanced for the observed trend: STOUT (1982) has reported a reduc.- ACKNOWLEDGEMENTS tion in invertebrate abundances in tropical streams, resulting from the impacts of scoring discharge and 1 am grateful to Mr. Okon 1. Okon (Department of Zoology, HYNES (1975) has cited studies showing that the Universit.y of Uyo, Nigeria) for field assistance. Engr. Darlong down-stream drift of invertebrates in rivers posi- Peters (Texas, USA) provided financial assistance and some laboratory equipment. tively correlates with flow rate while WELOAIME (1979) reviewed works suggest.ing wider dispersion Alanuscrif accepti par le Comité de rédaction le 25 juillet 1995

Rerr. Hydrobiol. trop. 27 (2) : XI-1.Z (1$9$). R. P. I

REFERENCES

ALAT.~LO (R. V.), 1981. - Problems in t.he measurement. of guidelines. Environmental Research Laboratory Office evenness in ecology. Oikos. 37 : 199-204. of Research and Development, U.S. Environr.nental Prot.ection Agency, Corvallis, Oregon 97331. 120 p. AL~.~RET (J. J.), 1982. - Reproduction et fécondité des pois- sons d’eau douce de Côte-d’Ivoire. Reo. Hydrobiol. trop., KING (R. P.), 1989. - Distribution, abundance, size and feed- 15 (1) : 347-371. ing habits of Brienomyrus brachyisfius (Gill, 1862) (Tele- ostei : ) in a Nigerian rainforest stream. AN~ERMEIER (P. L.), 1985. - Spatio-t,emporal patterns of Cybium, 13 (1) : 25-36. foraging success for fishes in a Illinois stream. Am. Midi. NU~., 114 (2) : 342-359. KING (R. P.). 1991. - The bioloou of Tilauia mariae Boufenoer, ~189.9 (Perciformes : Cichrdaé) in L Nigerian rain fo;es f ANGERMEIER (P. L.), KARR (J. R.). 1983. - Fish communit,ies sfream. Ph. D thesis, University of Port. Harcourt., Nige- along environment.aI gradients in a system of tropical ria, xiv + 237 p. st.reams. Enu. Biol. Fi&., 9 : 117-135. KING (R. P.), 1993. - Seasonal plasticity in faunal diet.ary BAGF.NAL (T. B.), 1978. - « Aspects of fish fecundit.yr. In GER- stat.us, diversity and foraging performance of Ilisha KINS (S. D.), éd. : Ecology of freshwater @sh produciion. africana (Bloch, 1795) (Clupeidae) in Qua Iboe est.uary, Blackwell Scient.ific Pub : 75-101. Nigeria. Cyhium, 17 (4) : 287-298. BAILEI. (N. T. ,J.), 1959. - Sfatisfical mefhods in biology. TIie English Language Book Societ.y and the English LJni- KOCH (L. P.), 1957. - Index of biotal dispersity. Ecology, 38 : vrrsities Press Ltd., 200 p. 145-148.

BERC; (J.), 1979. - Discussion of methods of investigat,ing the LOISELLE (P. V.), 1971. - The biology of Epipfafys sexfascia- food of fishes, with reference t.o a preliminary study of fus Gill, 1862 (.omi : Cyprinodontidae : Rivuli- the prey of Gobiusculus flaoescens (Gobiidae). Marine nae) in southeastern Dahomey and Togo hills. JzlK4, 7 Biology, 50 : 263-273. (2) : 29-53.

BOND (E. E.), 1979. - Biology of fishes. W.B. Saunders LOWE-MCCONNELL (R. H.), 1975. - Fish communifies in frop- Company. Philadelphia, 514 p. ical freshwafer : fheir distribution, ecology and eoolufion. Longman, London. 337 p. CI.AUSEN (H. S.). 1964. - Correlation of ichthyofaunal distri- bution with distribution of geological deposits and che- LOWE-MCCONNELL (R. H.), 1987. - Ecological sfudies in frop- mical composit.ion of freshwater in Nigeria. Videnak. ical fish communifies. Cambridge Univ. Press. 382 p. Medd. fra Danski Nalurh Foren Bd., 126 : 317-333. LOWENTIN (R. C.), 1966. - On t.he measurement of relative DA.JW (R.), 1977. - Jntroducfion to ecology. (Translated by variability. Sysfemaiic Zoology, 15 : 141-142. A. SOUTH), Hodder and Stought.on, London. 416 p. MAGALHAES (M. F.), 1992. - Feeding ecology of the Iberian DOUM (E. P.). 1971. - Fundamenfals of ecology. W.B. Saun- cyprinid Barbus bocagei Steindachner, 1865 in a lowland i~ers Company, 57-l p. river. J. Fish Biol., 40 : 13-133.

GRUNDEI. (R.). 1990. - The role of dietary diversity, prey Moss (B.) and EATON (J. W.), 1966. - The est.imation of num- capture sequence and individuality in prey selection by bers and pigment content in epipelic algae populations. parent mountain Chickadees (Parus gambrfi). J. Anim. Limnol. Oceanogr., 11 : 584-683. Ecol., 59 : 959-976. NIKOLSKY (G. V.), 1963. - The ecology of fishes. Academic HOLDEN (M.), REED (FI’.), 1972. - Wesf African freshwater ffsh. Press, London, 352 p. Longman, London, 68 p. PETER~ (R. H.), 1991. ‘- (ILesson from t.he size efficiency HYNE~ (.J. D.), 1975. - Dowstream drift of invert.ebrates in a river in sout.hern Ghana. Freshwaf. Biol., 5 : 515-532. hypot.hesis. 1. The general refuge concept.)) In LANZA- VECHHIA (G.) and VALVASSORI (R.),éd. : Form and fmc- K~RR (J. R.), TOTH (L. A.), GARMAN (G. D.), 1981. - Habifaf fion in Zoology. Selected Symposia and Monographs prPseroation for midwesf sfreom fishes : principles and U.Z.I., 5. Murchi, Modena : 335-361.

Rw. Hydrobiot. trop. &7 (2) : 1~3-3-1.5~~(1994). SEASONAL TROPHIC STATU8 OF P.AFER (NIGERIA) 155

REED (W.), BURCHARD (J.), HOPSON (A.J.), JENNES (J.), YARDWELCOMME (R. L.), 1985. - River fisheries. FAO Fish. Tech. (I.), 1967. - Fish and fisheries of Northern Nigeria. Pap., 262, 330 p. Ministry of Agricult.ure, Northern Nigeria, ‘226 p. WELCOMME (R. L.), 1986. - c

TEUGELS (G. G.), REID (G. Mc G.), KING (R.P.), 1992. - WHYTE (S. A.), 1975. - Dist,ribution, t.rophic relationships Fishes of the Cross River Basin (Cameroon-Nigeria) : and breeding habits of the fish populations in a tropical , zoogeography, Ecology and Conservat.ion. lake basin (Lake Bosumtwi-Ghana). J. ZOO~.Land., 117 Musée Royal de l’Afrique Centrale, Tervuren, Belgique. (1), : 25-56. Annales Skences Zoolo&ues, vol. 266, 132 p: - WOOTON (R. J.).II 1973. - The effect of size of food ration on THOMAS (3. D.), 1964. - Studies on the growth of trout., Salmo egg production in the female t.hree spined-st.ickleback trutfa from four cont.rastine habitats. Proc. 2001. Soc. Gasterosteus aculeafus L. J. Fish Biol. 5 : 89-96. Land., 142 (3) : 459-509. - ZARET (T. M.), RAND (A. S.), 1971. - Composit.ion in tropical WmcobrnrE (R. L.), 1979. - Fisheries ecology of floodplain stream flshes : support for the competit.ive exclusion rioers. Longman Group Ltd., London. 317 p. principle. Ecologg, 52 : 336-342.

R~U. Hydrobiol. trop. 27 (2j : 14.3455 (1994).